MLV were prepared using a standard laboratory scale method which involves dissolving and mixing the selected lipids in the required ratio (DOPC/DOPE/OA/Chol 2/0.05/1.95/1/1) in an organic solvent buffer of chloroform/methanol (4/1 v/v) and the subsequent removal of the solvent by way of rotary evaporation to form a “lipid cake” on the bottom of a round bottom flask. The film was then hydrated with agitation resulting in the formation of a suspension of MLVs, which was milky in appearance. This method of MLV preparation resulted in the formation of a suspension, which had a high level of heterogenicity in the size distributions of the particles formed as can be seen from the size distribution plot in Fig 3.2. This data shows that the particle mean size of the MLV suspension distribution was in the order of 500 nm and the polydispersity was of 0.4 or above indicative of a broad distribution and heterogeneous particle population.
In Fig 3.2 a size distribution plot for a suspension of liposomes after a single pass through the jet homogeniser is also presented at an operating pressure of 15 KPSI equivalent to
103.4 mNm'^. This distribution plot demonstrates that a single pass through the jet homogeniser at this operating pressure was sufficient to produce a suspension of SUV with a distribution having a mean particle size of approximately 140 nm and a polydispersity of 0.3.
Chapter 3 Leigh A Maguire. Fig 3.2. 1 0-, 4 - 2- 200 400 600 800 1000 Size (nm)
Typical size distribution profiles for a MLV suspension (■) prepared using the standard drying down and re-suspension method previously described and a suspension of SUV produced from a single pass through the jet homogeniser at an operational pressure of
The ability of the jet homogeniser to form SUV of a size within the range suitable for in vivo gene delivery offers advantages over conventional methods of MLV downsizing, which often require multiple passages or cycles. Fig 3.3 shows the mean particle size of a distribution of liposomes as function of the number of cycles of sonication. The MLV suspension was subjected to cycles of sonication at a frequency of 15 KHz for a period of t min followed by 2 min of rest. As can be seen a single cycle resulted in a suspension with a distribution having a mean similar to that obtained with the jet homogeniser. Additional cycles of sonication had little effect upon the mean particle size of the distribution but resulted in a gradual clarification of the suspension (see Figure 3.4). However, the sonicated suspensions obtained were noticeably more turbid than those obtained from homogenisation with 8 cycles of sonication being required to obtain the equivalent turbidity measurement at 400 nm obtained from a single pass through the jet homogeniser. MLV suspensions are by nature turbid having a milky appearance and the clarification of a liposome suspension is consistent with the replacement of MLV with SUV in a suspension (Templeton & Lasic, 1999). This would therefore suggest that sonication is a less efficient method of MLV disruption than the jet homogeniser.
Chapter 3 Leigh A Maguire. Fig 3.3. 2 0 0- ] 1 9 0 - 1 8 0 - 1 7 0 - 1 6 0 - 1 5 0 - 1 4 0 - 1 3 0 - 120- ^ 1 1 0- S 1 0 0 - (U 90 - N CO 8 0 - 7 0 - 6 0 - 5 0 - 4 0 - 3 0 - 2 0- 1 0- 0 3 4 5 6 Cycles of Sonication
The mean particle size of liposomes following successive cycles of sonication at a frequency of 15 KHz for intervals of Imin followed by 2 min rest. Error bars represent standard errors (n=5).
Fig 3.4. E c o o TD ■e H 0 . 6 - 0 . 4 - 8 0 2 4 6 C y c l e s
Turbidity of a SUV suspension generated by successive cycles of sonication (■) at an amplitude of 15 KHz for cycles of 1 min exposure and 2 min rest. Also shown is the turbidity of a SUV suspension generated following a single pass through the jet homogeniser (•) at a pressure of 103.4 mNm'^. The turbidity measurements were taken at a wavelength of 400 nm.
Chapter 3________________________________________________ Leigh A Maguire. Sterility along with the control of pyrogens and endotoxins are issues that are of utmost importance in the production of any drug, especially those intended for administration via systemic injection. Many of the methods commonly in use for sterilising biopharmaceutical products have proven unsuitable with liposomes and when DNA is also taken into account the number is reduced still further. y-Ray and heat sterilisation have both been shown to damage liposomes (Watwe & Bellare, 1995) while autoclaving has been reported to adversely affect DNA (Zuidam et al, 1993). Probably the most suitable method for the production of sterile liposomes will therefore be the use of sterile filtration combined with an aseptic technique. A noticeable advantage of the jet homogeniser is that even after just a single pass the resulting SUV suspension had a size distribution suitable for sterilisation with a 200 nm sterile filter medium without causing excessive loss of liposomes from the suspension (Fig 3.5).
Following filtration, some reduction in the polydispersity of the distribution was seen. The polydispersity was reduced from 0.3 to 0.2, which is notable from fig 3.5. There was a slight reduction in the mean particle size of the distribution, which would be consistent with the removal of the largest unilamellar vesicles from the suspension. However the narrow distribution of the suspension prior to filtration resulted in this lipid loss being low. Typically the recovery of lipid following filtration was in the order of 88 %.
Fig 3.5.
1 2-,
1 0-
100 200 300 400 500 600 700 800 900 1000
Size (nm)
Typical size distribution profiles of SUV suspensions generated by a single pass through the jet homogeniser at 103.4 mNm'^ prior to (■) and following a single pass through a 200 nm sterile syringe filter medium (•).
Chapter 3________________________________________________ Leigh A Maguire.
3.3. Control of Liposome Size.
The ability to control in a reproducible manner the size of the liposomes would be extremely useful. The reproducibility of sonication was poor and this has also been reported elsewhere (Tsai, 1999). Also increasing the number of cycles of sonication was shown to have little effect upon the overall size of the liposomes produced only aiding in the removal of MLVs from the suspension. In order to investigate the controllability and reproducibility of the jet homogeniser the size distributions of samples were taken after each pass for up to five passes, at operating pressures of 103.4 mNm"^ to 172.4 mNm'^. Typical size distributions for each operating pressure after 1, 3 and 5 passages through the jet homogeniser are shown in Fig 3.6-3.S.
Together these three figures demonstrate that it was possible to control the size of the resulting SUV between approximately 80 to 130 nm by altering either the number of passes through the device or the operating pressure used. The mean particle size derived from distributions obtained from three replicate experiments are shown in fig 3.9. The small standard errors demonstrate that the reproducibility of the jet homogeniser is good especially when compared to the results for a method such as sonication.
This is a powerful tool. It is likely that different cellular targets will require complexes with different sizes. For example as has been discussed in chapter 1 section 1.3, ex vivo
gene delivery has been shown to favour complexes of a larger size than in vivo gene delivery. The ability to control the size of the liposomes produced reproducibly will
enable one device to produce liposomes for a variety of targets rather than needing several different methods which will also be of value in validation of processes.
Chapter 3 Leigh A Maguire. Fig 3.6. 12-1 9 - 6- 3 - 150 200 100 250 Size (nm)
Typical size distribution profiles of SUV suspensions formed following a single pass (■), three (•) and five passes (À) through the jet homogeniser at an operating pressure of
103.4 mNm'^. All samples were filtered using a 200 nm sterile syringe filter prior to recording the size distributions.
Fig 3.7. 12 -1 9 - c 3 - 0 50 100 150 200 250 Size (nm)
Typical size distribution plots for SUV formed by downsizing MLV using the jet homogeniser at 137.5 mNm’^ for 1 pass (■) three passes (•) and 5 passes (A). Prior to recording size distributions the samples were filtered using a 200 nm filter.
Chapter 3 Leigh A Maguire. Fig 3.8. 1 2-, 9 - 6- 100 150 200 250 Size (nm)
Typical size distribution profiles obtained by downsizing MLV using the jet homogeniser at an operating pressure of 172.4 mNm'^ for a single pass (■) three passes (•) and five passes (A). Prior to recording size the samples were filtered using a 200 nm filter.
Fig 3.9. 140-1 135 130 1 2 5 - 120 115 1 1 0- 1 0 5 - 1 0 0- 9 5 - 9 0 - 8 5 - 8 0 - 75 P ass
Mean particle size of SUV suspensions generated by downsizing MLV using the jet homogeniser at operating pressures of 103.4 (■) 137.5 (•) and 172.4 mNm'^ ( A ) as a function of the number of passages through the device. Prior to recording the size distributions of the suspensions the samples were filtered using a 200 nm filter medium. Standard errors are for three replicate experiments.
Chapter 3________________________________________________ Leigh A Maguire.